
Replication package guide and data sources list for "The Effect of Fox News on Health Behavior During Covid-19"
Corresponding authors: 
- Dominik Hangartner: dominik.hangartner@gess.ethz.ch
- Matteo Pinna: matteo.pinna@gess.ethz.ch

Replication package guide
---------------------------------------
The replication package includes four folders: "do_files", containing the do files for the analysis; "R_file", containing the R file for the survey analysis; "data", containing the data files in .dta format (Stata); and "output", an empty folder were the do-files save the analysis when run. 

There are four do-files and one R-file: one that installs the SSC packages required for analysis, one that generates the county level analysis for year 2020, one that runs the analysis for 2019, one do-file that generates the zipcode level analysis and one R-file that generates the survey analysis. Each file is independent and can be run separately.

Each analysis file contains a preamble: it sets some basic settings, it prompts the user to set as directory the replication package folder, provides a list of packages to install (also available separately), loads the data from the "data" folder and creates a set of globals needed to run the analysis. 
After that, the user can find the code generating the graphs/tables of the paper in the files, following the instruction written in it: each chunk of code points to the set of figures/tables it generates, following the denomination adopted in the paper. E.g. "analysis_county_2020.do" first analysis' code-chunk contains the code generating "figure 1 panel a, figure 1 panel b, figure s6, figure s7, figure s13, table s4, table s5".

After running the preamble sections of the code in each file, the user can therefore find within the file the item they are looking for. For example looking for "figure 1 panel a" in "analysis_county_2020.do" will point to the chunk of code that, when run, generates the results/figure for that item.


Data sources
---------------------------------------
Publicly available data:

1) Data on overall usage of television: American Time Use Survey, available at https://www.bls.gov/tus/data.htm

2) Demographics: 2010 U.S. Census (https://www.census.gov/programs-surveys/decennial-census/data/datasets.2010.List_327707051.html)

3) Networks' transcripts: GDELT Television Comparer API accessible at https://api.gdeltproject.org/api/v2/summary/summary

4) COVID-19 cases and deaths: COVID-19 Dashboard published by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University. Accessible at https://coronavirus.jhu.edu/map.html

5) US Elections data: MIT Election Data, available at: https://electionlab.mit.edu/data and from Martin, G. J., and A. Yurukoglu. 2017. “Bias in cable news: Persuasion and polarization.” American Economic Review 107 (9): 2565–99.

6) Additional employment data: American Community Survey, available at https://www.census.gov/programs-surveys/acs

7) Emergency state declarations by county: National Association of Counties (NACo), available at https://ce.naco.org/?dset=COVID-19&ind=Emergency%20Declaration%20Types

8) Crosswalks: University of Missouri’s Geocorr tool, at http://mcdc.missouri.edu/applications/geocorr2018.html

---------------------------------------

Proprietary data:

9) Television viewership ratings: Nielsen Media Research TV ratings
	Contacts: Stanley Hui, stanley.hui@nielsen.com

10) Television channel positions: Nielsen Media Research FOCUS database
	Contacts: Stanley Hui, stanley.hui@nielsen.com

11) Mobility data: SafeGraph Social Distancing Metrics (https://docs.safegraph.com/docs/social-distancing-metrics)
	Contacts: https://www.safegraph.com/academics

12) Shopping data: Decadata transaction-level data on Covid-19-related purchases (http://decadata.io/#data)
	Contacts: Chris Antipa // O +1 415 742 0558 // M +1 415 812 5762 // Chris@DecaData.io // http://decadata.io/

13) Ideology and partisanship: Gallup Polling Social Series
	Contacts: Jerry Hansen, jerry_hansen@gallup.com

--------------------------------------

New data:

The survey measures contained in survey.dta, which can also be posted on Dataverse.


* Note: the proprietary data listed are not included in the package, future replicators can gain access to them using the contact information provided.